Pay for better risk-adjusted outcomes and let’s cut down on waste

April 19th, 2017 / By Norbert Goldfield, MD, Richard Fuller, MS

While Congress is finding it difficult to reach consensus on how to improve health care in the United States, the one thing we all can agree on is that ever-rising costs are the fundamental problem afflicting our healthcare system. Don Berwick, an acting CMS administrator under President Obama, recently opined that a number of states have experimented with ways to deliver care more efficiently.1 Yet, if one examines the states highlighted by Dr. Berwick as positive performers—California, Massachusetts and Arkansas—healthcare costs in these states continue to rise.2-4 Nationally, healthcare costs are moving to unsustainable levels projected to hit 20 percent of GDP by 2025, which some researchers consider an optimistic assessment.5,6

There are many sources of rising costs, such as the increased expense of new (and old) technology, cost of pharmaceuticals, ever-rising unit prices of individual health services such as hospital stays, and an aging population to name a few. Some policy proposals seek to mitigate the effect of these cost drivers on premiums, which may reduce some premium costs for younger people, but do nothing to help reduce premiums for older or sicker patients.7

There are healthcare policymakers who prioritize expanding coverage, including protecting coverage for 24 million individuals insured under the Affordable Care Act. Others focus, equally understandably, on the need to cut healthcare costs. What is needed is an approach that can strike a balance and address both objectives.

Let’s focus for the moment on Medicaid and consider one way the federal government and the states could work together. They should articulate and act on a simple message: Pay for better outcomes to help control costs. To do this effectively, the federal government together with the states must consider four system components of a pay for better outcomes approach. These four components—some controversial and cutting-edge to be sure—can improve delivery system efficiency, making it possible to consider coverage of more individuals:

1. Clinically credible risk adjustment that encompasses all populations. This would ensure that people who are disabled with major heart complications associated with diabetes, for example, are placed into a separate category from diabetics who are not as disabled from heart complications. The risk adjustment process must also recognize the impact of socioeconomic disparities and make sure that health care professionals/providers are not discriminated against because they are taking care of high-cost, high-need patients such as low income schizophrenics who have diabetes.

2. Capitation of all services using clinically credible risk adjustment as outlined above and described below in more detail. This includes payments made by the federal government to states for the Medicaid program.

3. We need to link financial incentives to improved outcomes (more details on this controversial idea below). As we’ve discussed a number of times in this blog, there are five outcomes that can be translated into dollars. Together they account for approximately 90 percent of outcomes that are potentially preventable:8

  • Potentially preventable admissions (e.g. diabetes out of control),
  • Potentially preventable readmissions (an infection necessitating a hospitalization occurring many days after discharge for a hernia operation),
  • Potentially preventable outpatient services (e.g. preventable MRIs for back pain or operations for back pain),
  • Potentially preventable complications (e.g. pneumonia occurring several days into a hospital stay for a patient hospitalized with a stroke),
  • Potentially preventable Emergency Room visits (such as a visit to the ER for a cold).

4. Financial incentives to decrease these potentially preventable events must be accompanied by regular and detailed reports that detail opportunities for improvement in these five types of potentially preventable events (collectively refer to as PPEs).

What are the controversies with this approach? It challenges the status quo in innovative ways. The federal government together with the states should move toward financing everyone covered under Medicaid through a per-person, per-month payment (also known as capitation). The country is rapidly moving to this type of payment system. The federal government and the states simply need to expedite it.

The second idea is that capitation payment must use a clinically credible risk adjustment system.9 It must be one that financially incentivizes the enrollment of the sickest people, unlike the current Medicare risk adjustment system that sees enrollees as a collection of distinct diseases rather than a more or less complex individual. An ideal risk adjustment classification system not only matches costs to needs but supports efforts to manage those needs.

The third new idea, one fraught with difficulty and opportunity (large sums of money involved—hundreds of billions of dollars), is to expeditiously transition federal government payments to states to payments based in part on a risk-adjusted capitated system. Currently, the federal government reimburses states for the Medicaid program essentially based on state Medicaid expenditures. The Federal government and States need to move federal government payment to states for Medicaid program expenditures slowly but surely toward more payments being based on a risk-adjusted capitation system.

First a word of background. In 2016 total Medicaid spending was $575.9 billion, which is 3.1 percent of gross domestic product (up from 1.0 percent in 1982). Federal funding accounts for 63.1 percent ($363.4 billion) of Medicaid spending (up from 52.9 percent in 1982). Federal funding for Medicaid is primarily determined by the product of: the state specific Federal Medical Assistance Percentage (FMAP) and the actual state “reasonable” spending for providing services to Medicaid enrollees (subject to an Upper Payment Limit that is the amount Medicare would have paid for similar services). As richer states can afford to support these programs more easily, the FMAP percentage has historically been set higher for states with lower per capita income to provide greater federal support. Actual state spending per Medicaid enrollee varies dramatically across states, in part due to readily explainable differences in input costs (i.e., lower labor costs in low income states) and in part due to unknown and poorly understood factors such as the relative efficiency of the delivery system in the state.  Yet across the states, it is clear that there are tremendous monetary opportunities to improve outcomes quality and reduce waste (e.g. rebalance the institutional bias in LTSS).10

If the United States is to address the Medicaid cost curve in a meaningful way, we should move expeditiously, but carefully, toward a risk-adjusted capitation system. Risk-adjusted capitation would transform the current Medicaid federal funding system from a negotiated contractual relationship with states to a uniform regulatory structure. It’s an approach that is based on lessons learned from the successful implementation of the Medicare inpatient prospective payment system (IPPS). Instead of an unsystematic approach to Medicare/Medicaid innovation via the Center for Medicare and Medicaid Innovation (CMMI), risk-adjusted capitation offers states a strong incentive to control Medicaid spending in innovative ways. It greatly reduces opportunities for financial manipulations to artificially increase federal payments to states. And importantly, it greatly reduces complex and burdensome oversight requirements. Granted, there are differences in benefits between states and significant differences in socioeconomic disparities, both of which will need to be addressed. Key to solving these challenging issues is careful research and a graduated movement of the current Medicaid financing system to a capitated approach.

If this third step is controversial, adjusting capitated payment for outcomes using PPEs, or a similar methodology that can easily translate outcomes into dollars, may be even more so. Yet this third step is critical to maintaining the Medicaid program while managing the growing costs of health care. The federal government could directly adjust payments to states based on their outcomes performance or provide financial incentives to states to implement state specific pay-for-outcomes payment systems for Medicaid providers in the state.  Paying for outcomes also provides the key to understanding which benefits lead to better outcomes and under which clinical circumstances. Paying for better outcomes directly and in a positive manner addresses the concerns of those that believe that Medicaid reform in general, and changes in in the current Medicaid financing system in particular, only lead to cuts in benefits, coverage or both. In addition, paying for outcomes can be an effective marketing slogan that both sides of the political aisle could at least consider.

Is there any experience with this pay for outcomes approach? Not surprisingly, states have a dramatic incentive to address the cost control challenge and, in fact, several states are implementing variations of this pay for outcomes approach, including Illinois, Maryland, Minnesota, New York and Pennsylvania. States are already saving millions and in the aggregate can save billions, thus preserving insurance coverage and satisfying fiscal conservatives.11

Skeptics might assert that the Medicaid data to support implementation of a pay for outcomes approach is not available today. In fact it is. The Centers for Medicare and Medicaid Services (CMS) Transformed Medicaid Statistical Information System (TMSIS)12 collects data from all 50 states and is available today. Is it perfect? Absolutely not. But as the 1982 implementation of DRGs demonstrated, as soon as financial incentives are brought to bear, the data gets better.

While the status of the American Health Care Act (ACHA) is unclear, the need to address payment reform—especially Medicaid—remains. Is the innovative approach outlined in this blog the perfect solution? No, it is not. There are other pressing issues, not the least of which is the skyrocketing cost of pharmaceuticals, but even this can be addressed at least in part by a pay for outcomes approach.13 If the federal government and the states focus on building a decent society, one which strives to provide healthcare coverage for all its citizens (as many leaders are claiming as their platform), while at the same time reducing waste and inappropriate care, then the time to pay for better outcomes is truly now.

Richard Fuller, MS, is an economist with 3M Clinical and Economic Research.

Norbert Goldfield, MD, is medical director for 3M Clinical and Economic Research.


  1. Berwick D. Obamacare can survive Trump. The New York Times. Published April 1, 2017.
  2. Freyer FJ. Premiums soar 21 percent for popular health plan. The Boston Globe. Published September 9, 2016.
  3. Petersen M, Levey N. California Obamacare rates to rise 13% in 2017, more than three times the increase of last two years. LA Times. Published July 19, 2016.
  4. Taylor C. Arkansas Health Insurers Ask For Huge 2017 Rate Hikes. The Arkansas Project. Published May 25, 2016.
  5. Congressional Budget Office. CONGRESSIONAL BUDGET OFFICE COST ESTIMATE: American Health Care Act.; 2017.
  6. GOP Health Bill in Shambles, House Commences Two-Week Break. The New York Times. Published April 6, 2017.
  7. Keehan SP, Stone DA, Poisal JA, et al. National Health Expenditure Projections, 2016–25: Price Increases, Aging Push Sector To 20 Percent Of Economy. Health Aff. 2017;36(3):553-563. doi:10.1377/hlthaff.2016.1627.
  8. Goldfield N, Kelly WP, Patel K. Potentially Preventable Events. Qual Manag Health Care. 2012;21(4):213-219. doi:10.1097/QMH.0b013e31826d1d3a.
  9. Hughes JS, Averill RF, Eisenhandler J, et al. Clinical Risk Groups (CRGs): a classification system for risk-adjusted capitation-based payment and health care management. Med Care. 2004;42(1):81-90. doi:10.1097/01.mlr.0000102367.93252.70.
  10. Erica L. Reaves, Musumeci M. Medicaid and Long-Term Services and Supports: A Primer.; 2015.
  11. Millwee B, Goldfield N, Averill R, Hughes J. Payment system reform: one state’s journey. J Ambul Care Manage. 2013;36(3):199-208. doi:10.1097/JAC.0b013e318295fb8f.
  12. Medicaid administrative data. MACPAC. 2017     
  13. Fuller RL, Goldfield N. Paying for On-Patent Pharmaceuticals: Limit Prices and the Emerging Role of a Pay for Outcomes Approach. J Ambul Care Manage. 39(2):143-149. doi:10.1097/JAC.0000000000000136.